Large share of trials may be unreliable

Investigations and statistical forensics suggest that in some medical fields at least about one‑quarter of published clinical trials are problematic — ranging from bad methods to possible fabrication. The piece argues for routine checks: compulsory data sharing, stronger registry enforcement, statistical forensics, and institutional audits to protect patients and evidence synthesis. — If a significant portion of clinical trials are unreliable, treatment guidelines, regulatory approvals and public trust in medicine are at risk, creating a need for policy and oversight changes.